Best Practices

Recommended setup

Start with simple report types

Begin with "Page: lifetime overview" or "Page: cumulative posts performance" before moving to time-based reports. These don't require date ranges and are less prone to issues.

Use separate flows for different report types

Create one data flow per report type rather than trying to combine them. This makes troubleshooting easier and keeps your destination cleaner. You can combine them later using Append transformations.

Test with data preview first

Always check the data preview before running the full flow. This shows you exactly what will be exported and catches missing-data issues early.

Append historical snapshots

For time-series analysis, set your flow to append rather than replace. This builds a historical record over time so you can track trends.

Data refresh and scheduling

Schedule daily for performance insights

"Page: performance insights" and "Page: follower gains trend" change daily. Schedule these flows to run once per day, preferably outside 9 AM–5 PM (avoid peak API usage).

Run weekly for demographic reports

Demographic data (follower/audience demographics) changes slowly. Weekly or bi-weekly runs are sufficient and reduce unnecessary API calls.

Use lifetime snapshot for reference

Run "Page: lifetime overview" once per week or month as a reference point. Changes slowly and useful for high-level health checks.

Performance optimization

Limit date ranges for large datasets

If pulling 1+ years of post or performance data, split it into quarters using date ranges. This prevents timeouts and keeps exports manageable.

Choose one demographic dimension per flow

For demographic reports, select one breakdown dimension per flow (e.g., geography OR job title, not both). This speeds up the API call and avoids data explosion.

Common pitfalls

Do

• Test manually before scheduling

• Use separate flows for different report types

• Check the data preview before running

• Append data to build historical records

• Schedule during off-peak hours

• Verify your LinkedIn permissions before troubleshooting

Don't

• Don't run the same flow multiple times in quick succession

• Don't try to combine multiple report types in one flow

• Don't ignore null values in the preview — they indicate missing data

• Don't export years of post data in a single run

• Don't change your LinkedIn role without re-authorizing Coupler.io

• Don't assume the exported data matches LinkedIn's real-time dashboard exactly (there's always a slight lag)